tod_training features.config.yaml options
The wiki page doesn't list the options for this yaml file. Where are these listed and read in? I'm having very mixed results.
The wiki page doesn't list the options for this yaml file. Where are these listed and read in? I'm having very mixed results.
TOD is currently unsupported and is seeing a big overhaul. Once the API is stable we will release it again (we plan around July/August 2011). I updated the wiki for the parameters we use internally so please trust that. Other features/descriptors are possible. For example, for SIFT you would use: detector_type: SIFT extractor_type: sequential descriptor_type: SIFT detector_params: max_features: 2000. min_features: 1000. extractor_params: octaves: 3.
Please start posting anonymously - your entry will be published after you log in or create a new account.
Asked: 2011-03-26 02:29:02 -0500
Seen: 124 times
Last updated: Jun 17 '11
What's the difference between OpenCV's BRIEF and rBRIEF in tod_*?
Purpose of pose estimations in tod_detecting?
Does tod_training in object_recognition require fiducial markers on training images?
Searching for TOD, ending up at Ecto
Physical setup for collecting tod_training data?
Why do test images for tod_detecting contain fiducial markers?